ETH Library
Taking the lymphatic route:
dendritic cell migration to draining
lymph nodes
Journal Article
Author(s):
Teijeira, Alvaro; Russo, Erica; Halin, Cornelia
Publication date:
2014-03
Permanent link:
https://doi.org/10.3929/ethz-b-000079467
Rights / license:
In Copyright - Non-Commercial Use Permitted
Originally published in:
Seminars in Immunopathology 36(2), https://doi.org/10.1007/s00281-013-0410-8
Funding acknowledgement:
138330 - Elucidating the impact of inflammation on lymphatic vessel function and on the induction of adaptive immunity (SNF)
This page was generated automatically upon download from the ETH Zurich Research Collection.
For more information, please consult the Terms of use.
REVIEW
Taking the lymphatic route: dendritic cell migration
to draining lymph nodes
Alvaro Teijeira & Erica Russo & Cornelia Halin
Received: 31 October 2013 /Accepted: 25 November 2013 / Published online: 9 January 2014
#
Springer-Verlag Berlin Heidelberg 2013
Abstract In contrast to leukocyte migration through blood
vessels, trafficking via lymphatic vessels (LVs) is much less
well characterized. An important cell type migrating via this
route is antigen-presenting dendritic cells (DCs), which are
key for the induction of protective immunity as well as for the
maintenance of immunological tolerance. In this review, we
will summarize and discuss current knowledge of the cellular
and molecular events that control DC migration from the skin
towards, into, and within LVs, followed by DC arrival and
migration in draining lymph nodes. Finally, we will discuss
potential strategies to therapeutically target this migratory step
to modulate immune responses.
Keywords Dendritic cells
.
Migration
.
Lymphatic vessels
.
Draining lymph node
.
CCL21
Introduction
Vaccination is considered one of the greatest public health
achievements of the twentieth century and has contributed to a
dramatic decline in mortality from infectious diseases.
Antigen-presenting dendritic cells (DCs) are an important
immune cell type that is activated during both vaccination
and infection. Upon encountering a pathogen or a vaccine in
peripheral tissues, DCs take up antigen, mature, and start to
migrate via afferent lymphatic vessels (LVs) to draining LNs
(dLNs), where they present the antigen to T cells for the
induction of adaptive immune responses. Since their discov-
ery approximately 40 years ago [1], ample evidence has
demonstrated the importance of DCs not only in the induction
of adaptive immunity in the context of vaccination and infec-
tion but also for the maintenance of tolerance [2–4]. The
recognition of the importance of DCs in the immune system
was also prominently highlighted by the awarding of the 2011
Nobel Prize of Physiology and Medicine to the late Ralph
Steinman “for his discovery of the dendritic cell and its role in
adaptive immunity” [5].
Although the migratory pattern of DCs from peripheral
tissues to draining LNs has been known for approximately
30 years, many details about the cellular and molecular events
that govern DC migration via afferent LVs are only now
starting to be unraveled. This is very much in contrast to
leukocyte trafficking through blood vessels (BVs), which
has been studied in greater detail over the past 30 years. The
gap in knowledge can likely be attributed to the fact that also
the lymphatic vascular system as a whole has been much less
well studied in comparison to the blood vascular system.
Although LVs were already described in the seventeenth
century, and the embryonic development of lymphatics was
extensively studied during the beginning of the twentieth
century [6], the lack of specific molecular markers for long
time hampered further investigation of the lymphatic vascular
system. In fact, lymphatic markers, such as the vascular en-
dothelial growth factor receptor-3 (VEGFR-3), the
hyaluronan receptor LYVE-1, podoplanin (gp38), or the
lymphatic-specific transcription factor Prox-1 have only been
identified over the past 18 years [6, 7]. Their discovery has
contributed to a true renaissance and explosion of lymphatic
vascular research and to recent progress made in the field of
leukocyte trafficking via LVs.
In this review, we will first give a brief introduction to LV
and DC biology. We will particularly focus on the skin and
skin-dLNs, as these are the organs in which DC migration via
This article is a contribution to the special issue on New paradigms in
leukocyte trafficking, lessons for therapeutics - Guest Editors: F. W.
Luscinskas and B. A. Imhof
A. Teijeira
:
E. Russo
:
C. Halin (*)
Institute of Pharmaceutical Sciences, Swiss Federal Institute of
Technology, ETH Zurich, Wolfgang-Pauli Str. 10, HCI H413,
8093 Zurich, Switzerland
e-mail: cornelia.halin@pharma.ethz.ch
Semin Immunopathol (2014) 36:261–274
DOI 10.1007/s00281-013-0410-8
LVs has mainly been studied. In a next step, we will introduce
some of the tools, which are commonly used to experimen-
tally study DC migration via LVs in vivo. We will then
summarize and discuss different aspects of DC migration in
the order of the DC’s physiologic itinerary, namely its migra-
tion from peripheral tissues towards, into, and within lymphat-
ic vessels (LVs), followed by its arrival and migration in the
dLN. Finally, we will discuss emerging experimental and
clinical approaches to modulate DC migration for the im-
provement of DC-based vaccines or for the prevention of
transplant rejection.
Characteristic features of afferent LVs
The lymphatic vascular system is essential for fluid drainage
from peripheral tissues, and for the uptake of dietary fats in the
intestine [8]. Moreover, it fulfills important immune functions
by mediating the transport of soluble antigens to dLNs as well
as leukocyte trafficking to and from LNs. With the exception
of the brain, LVs are present in virtually all vascularized
tissues. The afferent lymphatic network begins in form of
blind-ended lymphatic capillaries, which are considerably
wider than BVs. Lymphatic capillaries then merge into
collecting vessels, which may span long distances and finally
connect the afferent lymphatic network with a dLN (Fig. 1a).
Afferent LVs connect with the collagen-rich capsule of the LN
and drain lymph directly into the space below the capsule,
which is known as subcapsular sinus (SCS). The bottom of the
SCS is lined by LYVE-1
+
-positive lymphatic endothelial cells
(LECs), which are interspersed with CD169
+
macrophages.
After passing through the SCS, lymph flows into the surface-
rich and highly branched medullary sinus. In the paracortex, in
proximity to high endothelial venules (HEVs) and the T cell
area, blind-ended LYVE-1
+
cortical sinuses begin and repre-
sent the sites of lymphocyte egress from LNs [9]. Like the
SCS, also the cortical sinuses merge with the medullary sinus,
which finally converges into an efferent LV that exits the LN
[10](Fig.1a). Particularly in larger mammals, lymph is often
transported sequentially through many LNs that are organized
in chains. In this setup, an efferent LV can at the same time
represent the afferent LV of the subsequent LN [11]. In this
review, however, we will explicitly refer to the initial LVs in
peripheral tissue, which are composed of both capillaries and
collectors, when mentioning the term afferent LVs. After
passing through one or several LNs, the collecting LVs finally
merge in the thorax to form a single conduit named thoracic
duct, which releases its content (i.e., lymph) into the blood
vascular circulation at the level of the left subclavian vein.
The morphology of afferent LVs is ideally adapted to their
function, namely the uptake and transport of tissue fluids and
leukocytes: Blind-ended lymphatic capillaries are surrounded
by a very thin and highly fenestrated basement membrane,
which is composed of collagen IV, laminin, perlecan, and
nidogen [12, 13](Fig.1a). LECs in lymphatic capillaries have
a unique oak leaf shape [14]. Neighboring oak leaf-shaped
LECs partially overlap and are connected to each other by
discontinuous cell–cell junctions (Fig. 1b), which are arranged
in “button-like” associations of adhesion molecules. Such
buttons contain vascular endothelial cadherin and tight
junction-associated molecules such as claudin-5, occludin, or
junctional adhesion molecule-A [14]. This unique pattern of
tight junctions and partially overlapping LECs generates char-
acteristic flaps (also called primary valves), which are permis-
sive to the passage of fluids and macromolecules [14]. More-
over, the flaps are thought to be the prime site where leuko-
cytes enter into LVs [12, 14] (Fig. 1b). While lymphatic
capillaries are ideally suited for the uptake of lymph compo-
nents, lymphatic collectors are uniquely adapted to the trans-
port of lymph: Similarly to endothelial cells in BVs, LECs in
collecting LVs adopt an elongated shape and are surrounded
by a continuous lining of cell–cell junctions, which renders
the collecting LVs less fluid-permeable [14] (Fig. 1a, b).
Collecting LVs also contain specialized valves that impede
retrograde lymph flow and divide the vessel into segments,
which are called lymphangions (Fig. 1a). Compared to lym-
phatic capillaries, lymphatic collectors are surrounded by a
much thicker and less fenestrated layer of basement mem-
brane. Moreover, lymphatic collectors are covered by smooth
muscles cells, which account for the rhythmical contractions
of collecting vessels. These contractions mediate the propa-
gation of lymph from one lymphangion to the next, in down-
stream direction.
CCL21 expression
A molecule with key relevance for DC migration via LVs is
the chemokine CCL21, which is constitutively expressed by
LVs [15–17](Fig.2a, b). In response to activating stimuli,
tissue-resident DCs upregulate the CC-chemokine receptor 7
(CCR7) [2, 18], which initiates their migration towards
CCL21-expressing LVs and to dLNs. Indeed, in mice, block-
ade of CCL21 [15] or genetic deletion of CCR7 [19, 20]was
shown t o severely compromise DC migration to dLNs.
CCL21 comprises a highly positively charged C-terminal
motif, which accounts for its immobilization on heparan sul-
fates present on cell surfaces or in the extracellular matrix
(ECM) surrounding BVs and LVs [17, 21, 22]. Moreover,
CCL21 was shown to bind with nanomolar affinity to LEC-
expressed podoplanin, a mucin-type glycoprotein [23]. In
mice, CCL21 is encoded by two genes, which give rise to
two gene products: CCL21-Leu and CCL21-Ser differ in only
one amino acid but display gross differences in their tissue
distribution [24]. CCL21-Leu is the main isoform expressed
by LVs [24], whereas CCL21-Ser is mainly expressed by
HEVs and fibroblastic reticular cells (FRCs) in secondary
262 Semin Immunopathol (2014) 36:261–274
lymphoid organs (SLOs) [25]. Substantial insights into these
expression patterns and their functional consequences have
come from the analysis of a naturally occurring mutant mouse
strain, the so-called plt (paucity of lymph node T cells) mice.
plt mice have a defect in the production of CCL19 (the second
ligand of CCR7) and CCL21-Ser but retain expression of
CCL21-Leu in peripheral LVs [24, 26]. As a result, DCs in
plt mice are still able to enter into dermal afferent LVs, but
their entry and interstitial migration in the dLN are impaired
[25, 26].
Dendritic cells
DCs are important immune sentinels that form a functional
bridge between innate and adaptive immunity. Being particu-
larly abundant in tissues, which form the border with the
environment—such as the skin, intestinal tissues or the respi-
ratory tract—DCs readily come into contact with pathogens or
other noxious stimuli that breach the body’sbarriers.DCsare
highly phagocytic and derive their name from the fact that
they posses many dendritic process [1], which allow them to
constantly sample their environment. They express a plethora
of receptors that are able to recognize pathogen-associated
molecular patterns, including Toll-like receptors (TLRs) and
RIG and Nod-like receptors. Moreover, DCs can be indirectly
activated by recognition of damage-associated molecular pat-
terns or by inflammatory cytokines produced in the context of
tissue inflammation [2]. DC activation initiates a series of
phenotypic changes that are summarized as maturation [2].
During the maturation process, DCs cleave ingested antigen
into peptides for presentation on major histocompatibility
complex molecules and also upregulate co-stimulatory mole-
cules, which are essential for the subsequent activation of
naive CD4
+
and CD8
+
T cells. Besides enhancing the antigen
presentation capacity of DCs, maturation also induces pro-
found changes in the DC’s migratory behavior. Importantly,
DCs downregulate inflammatory chemokine receptors and
upregulate CCR7 a nd CXCR4 chemokine receptors [18].
While DC migration is strikingly enhanced in the presence
of infection and other forms of tissue inflammation, also a
minor but constant migration of DCs to LNs occurs in
uninflamed, steady-state conditions. In fact, steady-state DC
migration was shown to be important for the maintenance of
Fig. 1 Structure of afferent LVs and the LV network in dLNs. a Afferent
LVs begin as blind-ended capillaries, which merge into collecting vessels
and connect with dLNs. In contrast to lymphatic collectors, lymphatic
capillaries have a thin and a highly fenestrated basement membrane
(BM). Lymphatic collectors are surrounded by smooth muscle cells
(SMCs) and contain valves. A lymphangion ( LA ) is defined as the
segment between two valves. The lymphatic network in the dLN is
organized into the subcapsular sinus (SCS), the medullary sinus (MS),
and the cortical sinuses (CS). Further abbreviations used: T cell zone
(TCZ); B cell follicle (BF ), efferent lymphatic vessel (ELV ). b LECs in
lymphatic capillaries are oak leaf shaped and display a discontinuous,
“button-like” distribution of junctional adhesion molecules (red dots ).
Adjacent oak leaf-shaped LECs partially overlap, thereby creating open
flaps, which are also called primary valves. LECs in lymphatic collectors
have an elongated shape and are connected by continuous, “zipper-like”
cell junctions (red lines )
Semin Immunopathol (2014) 36:261–274 263
peripheral tolerance [27, 28]. In particular, mice that lack the
CCR7 receptor have been shown to develop symptoms of
autoimmunity in various organs [29].
DCs present in peripheral organs like the skin do not form a
homogenous population but exist in many different subsets of
functionally related cells [2, 4, 30]. In the skin, the organ in
which DC migration has been best studied, DCs can be very
rudimentarily divided into Langerhans cells (LCs), which are
found in the epidermis, and the various DC subsets present in
the dermis [31, 32](Fig.3a–c). LCs developmentally origi-
nate from the bone marrow (BM) but have the capacity to self-
renew in the epidermis under steady-state conditions [33].
However, under inflammatory conditions, new BM-derived
cells are recruited to the epidermis and differentiate into LCs
[34]. LCs are characterized by the expression of CD207/
langerin, epithelial cell adhesion molecule (EpCAM) and
CD1a [32]. Both in mice and in humans, the dermis is popu-
lated by various subsets of DCs that are referred to as dermal
DCs. Under steady-state conditions, murine dermal DCs are
replaced approximately every 10–14 days by precursors from
the BM [4]. In the context of inflammation, the dermal DC
pool is additionally increasedbyrecruitedinflammatory
monocytes that can differentiate into monocyte-derived DCs
[4]. Simply spoken, dermal DCs can be divided into two
subsets, depending on their expression of CD207/langerin
and CD103, or the expression of CD11b [4, 35]. All skin-
resident DC subsets can migrate to LNs, but the exact contri-
bution of the individual subsets to immune responses is not yet
completely understood [4, 35, 36].
Tools to study DC migration in vivo
Experiments to investigate the involvement of a gene product
in DC migration to dLNs rely on the availability of knock-out
mice or blocking antibodies. Traditionally, such experiments
have involved the adoptive transfer of DCs into recipient mice
[37–41], or the performance of so-called fluorescein isothio-
cyanate (FITC) painting experiments [16, 41–44]. In adoptive
transfer studies, DCs are typically isolated from donor mice or
generated in vitro from BM cultures. Upon fluorescent or
radioactive labeling, DCs are injected into the skin, e.g., the
footpad of a recipient mouse, and their arrival in the dLNs is
analyzed and quantified 16–48hlater.InFITCpainting
experiments, on the other hand, FITC dissolved in dibutyl
phthalate is applied onto the skin of mice. FITC is a contact
sensitizer and rapidly penetrates the skin, leading to its uptake
by skin-resident LCs and dermal DCs. The painting process
also induces DC mobilization to dLNs, where DCs are subse-
quently quantified based on their green fluorescent signal.
More recently, also confocal- and multiphoton-based time-
lapse microscopy [45] has started to become a useful tool for
elucidating DC migration via LVs. In contrast to the above-
mentioned migration experiments, which address the overall
involvement of a molecule in the migratory process, imaging
studies allow the investigation of DC migration in situ and
with cellular resolution. Consequently, imaging experiments
identify the step in the migratory process a particular candi-
date gene participates in, i.e., whether this is in the migration
towards, into, or within afferent LVs, or migration within
Fig. 2 Expression of CCL21 in
LVs. The expression of CCL21
(green) in LYVE-1
+
LVs (red )
was analyzed in tissue whole
mounts prepared from murine ear
skin. a Staining of the
extracellular fraction of CCL21
(performed under unfixed
conditions, as described in [17])
reveals a diffuse CCL21 staining
pattern that largely co-localizes
with LYVE-1
+
LVs. b Staining
performed under PFA-fixed and
permeabilizing conditions (as
described in [16]) suggests the
presence of punctuate,
intracellular deposits of CCL21 in
LECs. Scale bar,100μm
264 Semin Immunopathol (2014) 36:261–274
![Fig. 2 Expression of CCL21 in LVs. The expression of CCL21 (green) in LYVE-1+ LVs (red) was analyzed in tissue whole mounts prepared from murine ear skin. a Staining of the extracellular fraction of CCL21 (performed under unfixed conditions, as described in [17]) reveals a diffuse CCL21 staining pattern that largely co-localizes with LYVE-1+ LVs. b Staining performed under PFA-fixed and permeabilizing conditions (as described in [16]) suggests the presence of punctuate, intracellular deposits of CCL21 in LECs. Scale bar, 100 μm](/figures/fig-2-expression-of-ccl21-in-lvs-the-expression-of-ccl21-2l48brzs.webp)